Reliable, efficient, wavelength-agile, high-brightness, room temperature laser sources operating in the mid-infrared (IR) and terahertz (THz) region are in great demand for a wide variety of applications in defense (IR countermeasures, laser radar, IR communications); security (scanners, remote sensing of chemical and biological agents); industry (gas sensing, leak detection, pollution monitoring, process control); science (THz spectroscopy); and medicine (IR and THz medical images). Over the years the search for a solution has considered a number of different approaches, starting with the investigation of a great variety of binary and ternary semiconductor materials.
One of the most promising approaches is to convert the frequency of available mature pump lasers into the waveband of interest via a nonlinear material. Initially employing bulk birefringent phase-matching (BPM) crystals, the idea was eventually extended to exploit quasi-phase matched (QPM) materials. The first such practical structure was realized in periodically-poled lithium niobate (LiNbO3) (PPLN), but strong intrinsic absorption has limited the usage of this and similar ferroelectric crystals to wavelengths shorter than about 4 μm, i.e. they are partly usable for only the first of two atmospheric transparency windows (2 μm to 5 μm, but not 8 μm to 12 μm).
In non-ferroelectric materials, quasi-phase matching can be achieved by spatially inverting the nonlinear susceptibility. Initially this was done through alternating the orientation of thin wafers ordered in a stack. However, the high optical losses due to the multiple wafer interfaces, the small thickness required, as well as the tight fabrication tolerances made this approach inefficient for practical devices. Quasi-phase matching is now more favorably realized in semiconductors by growing thick layers of material on thin templates, which have been designed in advance with the desired pattern. For example, OP-gallium arsenide (OPGaAs) fabricated using this approach has emerged as a promising nonlinear material, due to its broad IR transparency, high nonlinear optical susceptibility, and ability to be grown on OP-templates following the engineered periodic structure.
However, significant two-photon absorption (2 PA) in the wavelength range of about 1 μm to about 1.7 μm has limited QPM GaAs from taking advantage of many convenient pump laser sources such as Nd-, Yb- and Cr-doped yttrium aluminum garnet (YAG) lasers radiating in this region, or laser diodes and Er-doped fiber lasers radiating around the important telecommunications wavelength 1.55 μm. There have been isolated attempts in industry to work around these pump source issues rather than solve them. For example, one of them uses a 1-μm-pumped optical parametric oscillator (OPO) to provide output from 1.7 μm to 2 μm, which is beyond of the high end of the GaAs 2 PA range. This output wavelength can then be used as a pump for the GaAs QPM structure. The inclusion of a second nonlinear conversion stage, however, increases the complexity of the overall system as well as the cost of the final product. Accordingly, the high 2 PA of GaAs at 1 μm to 1.7 μm is a serious limitation that has naturally led to interests in other materials, and particularly gallium phosphide (GaP) as a potential QPM material.
In contrast with GaAs, GaP has a negligible two-photon absorption (2 PA) in the 1 μm to 1.7 μm wavelength range, and a comparable nonlinear susceptibility. In addition, GaP has twice the thermal conductivity (at much lower thermal expansion coefficients) and also broad transparency, which, in contrast with GaAs, conveniently starts in the visible portion of the spectrum facilitating alignment in an optical setup. Due to these factors, interest in QPM GaP has increased rapidly, and frequency conversion devices (FCDs) based on stacks of GaP wafers have already been designed. The epitaxial growth of periodic GaP structures by Molecular Beam Epitaxy (MBE) and metalorganic chemical vapor deposition (MOCVD) has also been initiated. Hydride vapor phase epitaxy (HVPE), however, appears to be the only approach that is capable of producing apertures large enough for high power applications.
In order to produce high quality OP materials, high quality (e.g., minimal defects) OP templates are desired. Unfortunately, OPGaP templates having sufficiently high quality suitable for homoepitaxial growth are presently unavailable, partly due to the low quality (e.g., high etch pit density (EPD) and poor parallelism) of the commercially-available GaP wafers, and partly due to the not-yet-optimized template preparation techniques. Thus, the OPGaP material grown by HVPE on the available OPGaP templates is generally of low quality.
Accordingly, there is a need for high quality OPGaP materials and methods for making them.